Most of the sanitary hot water used in South African buildings is heated by means of direct electrical resistance heaters. This is one of the major contributors to the undesirably high morning and afternoon peaks imposed on the national electricity supply grid. For this reason, water heating continues to be of concern to the electricity supplier, ESKOM. Previous studies, conducted by the Potchefstroom University for Christian Higher Education in South Africa, indicated that extensive application of the so called inline heat pump water heating methodology in commercial buildings could result in significant demand side management savings to ESKOM. Furthermore, impressive paybacks can be obtained by building owners who choose to implement the design methodology on existing or new systems. Currently, a few examples exist where the design methodology has been successfully implemented. These installations are monitored with a fully web centric monitoring system that allows 24 h access to data from each installation. Based on these preliminary results, a total peak demand reduction of 108 MW can be achieved, which represents 18% of the peak load reduction target set by ESKOM until the year 2015. This represents an avoided cost of approximately MR324 (ZAR) [Int J Energy Res 25(4) (1999) 2000]. Results based on actual data from the monitored installations shows a significant peak demand reduction for each installation. In one installation, a hotel with an occupancy of 220 people, the peak demand contribution of the hot water installation was reduced by 86%, realizing a 36% reduction in peak demand for the whole building. The savings incurred by the building owner also included significantenergy consumption savings due to the superior energy efficiency of the heat pump water heater. The combined savings result in a conservatively calculated straight payback period of 12.5 months, with an internal rate of return of 98%. The actual cost of water heating is studied by comparing the cost of supplying 1 kWh of thermal energy by either a conventional electrical resistance heater or a heat pump heater. The cost for one of the installations, based on the cost of kilowatt-hour consumption as well as peak demand contribution, showed a decrease in cost from 35.4 c/kWh (ZAR) to 7.4 c/kWh (ZAR). This paper presents data from these actual installations, where the issues of demand side management and energy cost reduction are addressed. (C) 2003 Elsevier Ltd. All rights reserved.